i Programme: Resource recovery
English title: INTENSIFICATION OF LIGNOCELLULOSIC BIOETHANOL PRODUCTION PROCESS USING MULTI-STAGED MEMBRANE BIOREACTORS
Year of publication: 2019 Authors: Clarisse Uwineza
Supervisor: Amir Mahboubi Soufiani Examiner: Prof. Mohammad J. Taherzadeh
ii
Abstract
iii
Publications
A. Mahboubi, C. Uwineza, W. Doyen, H. De Wever, M.J. Taherzadeh (2019). Intensification of lignocellulosic bioethanol production process using continuous double-staged immersed
iv TABLE OF CONTENT
Abstract ... ii
Publications ... iii
1. Introduction ...1
1.1 Background and Problem description ...2
1.2 Purpose of the research work and its limitations ...7
2. Ethical and social aspects ...9
3. Materials and methods... 10
3.1 Pretreated wheat straw slurry ... 10
3.2 Inoculum and culture preparation ... 10
3.3 Yeast cultivation in semi synthetic media ... 11
3.4 Wheat straw slurry solids and ash measurements ... 11
3.5 Enzymatic hydrolysis of the wheat straw slurry ... 12
3.6 Enzymatic hydrolysis and yeast cultivation in shake flasks ... 13
3.7 Membrane bioreactor design, set-up and defining filtration parameters... 14
3.7.1 Membrane Bioreactor set-up ... 14
3.7.2 Pure water, wheat straw slurry and hydrolysate filterability measurements ... 15
3.7.3 Yeast cultivation on semi-synthetic media and wheat straw hydrolysate to determine optimal operating condition ... 17
3.8 Double stage membrane bioreactors for continuous enzymatic hydrolysis and fermentation... 18
3.8.1 Pure water filterability to define filtration parameters ... 18
3.8.2 Wheat straw slurry dilution and MBR assisted enzymatic hydrolysis in batch and continuous process ... 18
3.8.3 Yeast pre-culture medium preparation ... 19
3.8.4 MBR assisted fermentation batch process to optimize operating condition ... 19
v
3.9 Analytical methods ... 22
4. Results and discussion ... 24
4.1 Ethanol production in semi-synthetic media ... 24
4.2 Enzymatic hydrolysis of the wheat straw slurry, liquid and solid fractions ... 25
4.3 SSF and SHF enzymatic hydrolysis and yeast cultivation... 30
4.4 Effect of yeast cultivation on semi-synthetic media and wheat straw hydrolysate ... 32
4.5 Optimization of membrane filtration parameters ... 34
4.5.1 Comparison of clean water and hydrolysate filterability... 34
4.5.2 Comparison of constant fluxes to optimize interconnected membrane filtration performance ... 36
4.5.3 Yeast growth optimization for batch and continuous fermentation ... 38
4.5.4 Ethanol evaporation measurement in MBR ... 40
4.6 Double-stage membrane biorector ... 41
4.6.1 Double-staged submerged MBR performance at constant flux of 21.9LMH ... 43
4.6.2 Double-staged submerged MBR performance at constant flux of 36.4LMH ... 45
4.6.3 Double-staged submerged MBR performance at constant flux of 51 LMH... 47
4.7 Continuous enzymatic hydrolysis and filtration of pretreated wheat straw slurry .... 49
4.7.1 Continuous hydrolysis with low enzyme loading at constant permeate flux of 21.9LMH ... 49
4.7.2 Continuous hydrolysis with high enzyme loading at constant permeate flux of 21.9 LMH... 52
4.7.3 Continuous hydrolysis at constant permeate flux of 36.4 LMH ... 54
4.7.4 Continuous hydrolysis at constant permeate flux of 51LMH ... 56
4.8 Continuous ethanol fermentation and removal by submerged membrane bioreactor 58 4.8.1 Continuous cultivation with low inoculum at constant flux of 21.9 LMH... 58
4.8.2 Continuous cultivation with high inoculum at constant permeate flux of 21.9 LMH ... 60
4.8.3 Continuous cultivation at constant permeate flux of 36.4 LMH ... 61
vi
4.9 The effect of continuous membrane filtration for lignin recovery ... 66
4.10 The effect of medium, backwashing and permeate flux on filterability during hydrolysis ... 68
4.10.1 The effect of medium ... 69
4.10.2 The effect of flux ... 71
4.10.3 The effect of backwashing ... 72
4.10.4 Effect of initial suspended solids and medium viscosity. ... 73
Conclusion ... 76
Future prospective ... 77
Acknowledgment ... 77
1
1. Introduction
The development of the industrialized world today has the challenge in the limit of (often insufficient) energy sources which has resulted in an increase in fossil fuel utilization. As a consequence, the world is facing several problems including high fuel prices followed by the depletion of fossil fuels, increase in air pollution, high amount of greenhouse gases (GHG) emissions which exacerbate the problems of global warming and climate change. The continuous depletion and need for fossil fuels have encouraged considerable interest in the development of alternatives, renewable energy to fulfill today’s energy demands (Meihui et al. 2015). As the solution, bio based energy could be a sustainable alternative to usual fossil fuel based energy. As the research advances, more and more materials are being tested for the application sustainable production of biofuels such as bioethanol essential in transportation.
2 1.1 Background and Problem description
Lignocellulosic materials are considered as promising alternative source for second generation bioethanol production a renewable and environmentally friendly fuel. Up to date, numerous studies on second generation bioethanol technology have been done either at laboratory or pilot scale showing different demonstration on various lignocellulose substrates but still, actual bioethanol production is not yet commercially feasible at industrial scale (Lennartsson et al. 2014). The lignocellulosic materials are structurally composed of three main components: cellulose, hemicellulose, and lignin. Cellulose has a strong crystalline structure composed of polymer chains of carbohydrates of glucose attached together by strong bonds of beta-1,4 glycosidic. Hemicellulose is also mostly composed of polysaccharides and polyuronides rich in pentoses especially xylose and arabinose. Lignin as a relatively hydrophobic material is built by cross-linked aromatic polymer covalently bound together with hemicelluloses, forming a complex matrix that surrounds the cellulose micro-fibril by hydrogen bonds. The structures and compositions of these biopolymers vary greatly depending on plant species, geographical origin and growth conditions (Saini et al. 2014).
The recalcitrant structure of lignocellulosic materials is the greatest challenge in the production processes of second generation bioethanol. The tightness and complexity of this structure make them difficult for direct uses and the resistance to degradation comes from the high crystallinity of cellulose, the hydrophobicity of lignin and encapsulation of cellulose by the lignin-hemicellulose matrix (J.Taherzadeh et al. 2007). Therefore, the pretreatment of lignocellulosic aims to open up the complex structure of the cellulose-hemicellulose-lignin matrix, to increase the porosity of the structure and to enhance the accessibility and biodegradability of the polymer chains of carbohydrates of hexoses and pentoses for enzymatic hydrolysis (Govumoni et al. 2013; Taherzadeh et al. 2008).
A variety of pretreatment processes (fungal, irradiation, extrusion, alkali, acid, ozonolysis, organosolv, ionic liquids, steam explosion, liquid hot water, ammonia fiber explosion, wet oxidation, microwave, ultrasound, and CO2 explosion pretreatment) have been previously
3
inhibitors derived from pentoses and hexoses degradations respectively and carboxylic acids (such as acetic acid) from also hemicellulose, while a lot of phenolic compounds derived from degradation of lignin (Klinke et al. 2004; Taherzadeh et al. 2008). The type and amount of inhibitors generated depend on the lignocellulosic materials and the pretreatment methods used. Consequently, the presence of inhibitors can inhibit the selectivity of the fermentation process by decreasing cell growth by increasing longer lag phase, decreasing intracellular pH, preventing bioconversion of catabolic enzymes and disturbing cell membranes integrity, reducing volumetric ethanol productivity etc. (Taherzadeh et al. 2008). Furthermore, lignocellulosic pretreatment followed by Hydrolysis (acid or enzymatic) results in different monomeric sugars of hexoses (glucose, mannose etc.) and pentoses (xylose, arabinoses etc.) released from cellulose and hemicellulose respectively and can biologically be converted into bioethanol through fermentation (Taherzadeh et al. 2008).
In research done by J.Taherzadeh et al. (2007), enzymatic hydrolysis of lignocellulosic materials carried out by the help of enzyme cellulase has been reported as a promising process for cellulose hydrolysis compared to acid hydrolysis. Acid hydrolysis requires higher temperature and lower pH that forms a corrosive condition and yields more inhibitory compounds. On the other hand, enzymatic hydrolysis requires mild conditions and has the possibility to get to a high yield of cellulose hydrolysis. The main drawbacks of enzymatic hydrolysis are the slow nature of the process, the high price of the production of the enzymes and enzyme inhibition by the final product (high concentrations of sugars) (J.Taherzadeh et al. 2007). In order to overcome these issues, different strategies such as Simultaneous Saccharification and Fermentation (SSF) has been developed where the cells immediately consume sugars released by enzymes. However, SSF processes usually operate at suboptimal conditions since Cellulase enzyme and the yeast have different optimum conditions (Cellulase: 40-50 pH: 4.5-5.0 and yeast S. cerevisiae: 30-37). Separate hydrolysis and fermentation (SHF) remediate the issues with SSF (Ishola 2014).
4
industrial production of lignocellulosic generation bioethanol (Mahboubi et al. 2017a). Therefore, various approaches (Mahboubi et al. 2017a; Saleh A.A. 2008; Taherzadeh et al. 2007) have been approved that genetically engineered recombinant of S. cerevisiae has a high capacity to consume and ferment both glucose and xylose and enhance in situ detoxification of inhibitor at high cell concentration. In addition, lignocellulosic fermentation in either batch or continuous process has the main challenge of low ethanol yield and productivity due to the high concentration of inhibitors, residues mostly rich in lignin and high bacterial contamination which increase the production cost (Taherzadeh et al. 2007). The continuous cultivation is the most promising for fermentation of lignocellulosic hydrolysates. Compared to batch, it requires smaller investment; the reactor is all the time productive and has higher productivity. However, the presence of inhibitors in medium can prevent and limit the specific growth rate of the cells resulting in cell wash out of the bioreactor and a very low productivity (Taherzadeh et al. 2007). Figure 1 below shows the main issues related to the conventional bioethanol production process from lignocellulosic material stated above and in the following paragraph.
5
In order to have continuous great bioconversion of inhibitor, as well as preventing cell washout, a suitable mode of operation is necessary in the design of the process. According to Mahboubi et al. (2017a), to have complete sugars (xylose and glucose) utilization in the continuous fermentation process, higher capacity of inhibitory tolerance, and high rate of sugars conversion; substantial capacity of maintaining high cell concentration in bioreactor has to be promised.
Numerous studies (Mahboubi et al. 2017a; Westman et al. 2012; Ylitervo et al. 2014b; Ylitervo et al. 2011) has been done for the purpose of maintaining high cell concentration inside the reactor and cell recycling and utilization of different sugars such as cell immobilization through encapsulation and flocculation and membranes cell retention and reuse. The application of Membrane bioreactor in production of lignocellulosic ethanol has various advantages such as highest potential for high cell retention, ability of complete utilization of fermentable sugars (glucose and xylose), the capability of in situ detoxification of the bio-convertible inhibitors and avoidance of cell washout (Ishola et al. 2015; Mahboubi et al. 2017a; Mahboubi et al. 2016; Ylitervo et al. 2014a). Furthermore, Membrane Bioreactors (MBRs) has the high capacity to perform in continuous cultivation at higher dilution rate and low hydraulic retention time through the accumulation of cells inside the reactor. However, the use of high solid loading and viscous feed streams containing high suspended solids (mixture of sugars: pentoses, hexoses etc.) and the inhibitory compounds (Furfural, 5-hydroxymethylfurfural (HMF), carboxylic acids ) in fermentation media are still a great challenge in second generation bioethanol production process (Klinke et al. 2004).
6
membrane can separate enzymes and cells from liquid products depending on the type of membrane used Ultrafiltration, Reverse Osmosis, Nanofiltration or Microfiltration membranes. This project is aimed to intensify the second generation bioethanol production process from the ordinary process (figure 1) that has been used before, by the integration of submerged membrane bioreactors (MBRs) during hydrolysis and fermentation. In addition, the consumption of glucose and xylose, suspended solids control, high concentration of yeast inside the bioreactor, particle-free ethanol stream recovery in a continuous process, high productivity were investigated using pre-treated wheat straw slurry as substrates.
Table 1. Comparison ethanol yield from different process in second generation bioethanol using different microorganisms and different culture conditions.
Substrates Microorganisms Process conditions
Sugar content (g/l sugars)
Ethanol yield Reference Wood hydrolysate Saccharomyces cerevisiae Continuous cultivation Glucose 4.57 ±0.13 Xylose 3.21 ±0.13 0.44 ±.002g ethanol/g sugars (Ylitervo et al. 2014a) Wheat straw hydrolysate Recombinant Saccharomyces cerevisiae Anaerobic Continuous Fermentation Glucose 50 Xylose 50 0.42g ethanol/g sugars (Mahboubi et al. 2017a) Woody biomass Recombinant Saccharomyces cerevisiae Anaerobic Batch fermentation Glucose 5 Xylose 15 0.30g ethanol/g sugars (Saleh A.A. 2008)
Wheat straw
genetically-engineered strain of Saccharomyces cerevisiae Anaerobic batch fermentation Glucose 6 Xylose 21 30.3g/l equivalent to 83% of ethanol theoretical yield) (Ishola et al. 2015)
Wheat straw Saccharomyces
cerevisiae, Pichia stipitis and
co-culture of both Anaerobic fermentation Glucose 30 0.48 gp/gs, 0.43 gp/gs, and 0.40 gp/gs (Singh et al. 2012)
Wheat straw Saccharomyces
7
1.2 Purpose of the research work and its limitations
The purpose of this project was to intensify the lignocellulosic bioethanol production process from the ordinary process (figure 1) that has been used before, by the integration of submerged membrane bioreactors (MBRs) during hydrolysis and fermentation. The conventional production processes of second generation bioethanol as described above and in figure 1 suffers from low yields and productivity, have energy and cost intensive upstream and downstream processing and the feed substrate contains high amounts of suspended solid, inhibitory compounds and prioritized sugars. Integration of MBRs with bioethanol production process can be the key in tackling these issues.
In this project multi-staged integrated microfiltration membrane (figure 2) was applied in order to have continuous hydrolysis and fermentation using a recombinant xylose-consuming S. cerevisiae. Various value-added product streams of bioethanol, cell biomass and lignin from wheat straw slurry were produced. Separation of suspended solids, retaining enzymes, having high concentration of yeast inside the bioreactor, having a particle-free ethanol stream in a continuous high productivity process were the main targets in this approach. Different main goals are described below:
Application of MBRs in enzymatic hydrolysis for enzyme retention.
Fermentation of lignocellulosic substrates in continuous cultivations using membrane bioreactor
Preparation of interconnected automatically controlled double-stage MBR (the integration of continuous enzymatic hydrolysis using and continuous fermentation using microfiltration flat sheet submerged membrane)
Pretreated wheat straw slurry (Mahboubi et al. 2017a) were used as substrates in this project (this project was not focused on pretreatment).
To make this project thesis realistic and achieving high quality result different limitations has been recognized:
High concentration of suspended solids present in wheat straw hydrolysate is problematic in cell-suspended solid separation and membrane filtration.
8
To achieve the complete (glucose and xylose) into bioethanol, the reduction in contamination risks and product inhibitions were taken into account.
High solids loading was used to combat bacterial contamination, improve enzymatic hydrolysis and process efficiency.
The figure below describes the flowchart of the intensification of the second generation bioethanol production process using multi-staged membrane bioreactors (MBR):
9
2. Ethical and social aspects
10
3. Materials and methods
The experiments took place in different optimum conditions batch (for inoculum cultivation) and continuous in membrane bioreactors. Several processes have been examined in order to set-up interconnected new automatically controlled double-stage MBR. The pretreated wheat straw slurry was used as substrate. Enzymatic hydrolysis and fermentation for both batch and continuous were tested at different conditions such as different enzyme loading rates and temperatures. Cellulase Cellic Ctec2 was used in enzymatic hydrolysis; while recombinant Saccharomyces cerevisiae strain was used as active microorganism for fermentation. Semi-synthetic media (glucose, xylose, yeast extract, peptone and salt solution) were also used for experiments screening. Different analytical devices such as HPLC and spectrophotometer were employed during this research work. The projects experiments were categorized into six main stages in which each stage various test was targeted to be investigated: (1) Yeast cultivation and Inoculum preparation with semi-synthetic media; (2) Enzymatic Hydrolysis; (3) Membrane Filtration; (4) Continuous MBR assisted hydrolysis; (5) Continuous MBR assisted fermentation; (6) Interconnected double-stage MBR hydrolysis and fermentation.
3.1 Pretreated wheat straw slurry
The lignocellulosic substrate used in this project work is a Swedish agriculture biomass wheat straw pretreated in dilute-acidic condition (0.3-0.5%H2SO4 for 8 min at 185°C) by SEKAB
E-Technology (Örnsköldsvik, Sweden). After pretreatment, the wheat straw slurry was kept in a cold room (4-5) for further uses. Wheat straw has a high potential as sustainable biomass source in Europe based on its abundance and low cost (Mahboubi et al. 2017b).
3.2 Inoculum and culture preparation
11
20 g/l agar, 10 g/l glucose, 10 g/l xylose, 10 g/l yeast extract and 10 g/l peptone were measured with analytical balance and added to the 500 ml blue cup flasks containing 500 ml of Mill-Q water. The solution in the blue cup flasks was autoclaved for 20 min at 121°C. After sterilization the YPD solution was poured on 25 plates in aseptic condition. The plates were inoculated by the colonies of recombinant Saccharomyces cerevisiae stored in the fridge at 4 degrees and incubated for 2-3 days at 30°C. The plate’s lid was surrounded by parafilm and stored in the fridge for later uses. 3.3 Yeast cultivation in semi synthetic media
In order to prepare required yeast inoculum for cell growth optimization, yeast precultures were prepared in loop inoculated 250 ml Erlenmeyer flasks containing of yeast extract peptone dextrose (YPD) broth comprising 20 g/l Glucose, 10 g/xylose, 5 g/l peptone 5 g/l yeast extract and placed in a shaking water bath at 30°C and 120 rpm for 48 hrs.
The preparation of YPD broth was performed according to Mahboubi et al. (2017b) as follow: 800 ml of distilled water was added to a 1000 ml beaker placed on the magnetic stirrer to have a complete mixing. 20 g/l Glucose, 10 g/xylose, 5 g/l peptone 5 g/l yeast extract were weighted using analytical balance and added to the 1000 ml beaker containing 800 ml of distilled water. 100 ml of the mixed solution was added in each of six Erlenmeyer flasks of 250 ml lidded by cotton plugs and covered by aluminum foil and then autoclaved for 20 min at 121°C. The six flasks were then loop inoculated. After inoculation, the flasks were incubated in a shaking water bath running at 30°C and 120 rpm for 48h. The initial pH of the broth was measured to be around 5 after every 24 hours.
Sampling was performed in duplicate in the aseptic condition for 6, 24 and 48 h by taking 1 ml into micropipette tubes from each of the six flasks, centrifuged for 2 min at 15000×g and then the supernatant liquid was kept in the fridge for future analysis such as metabolites production. After 48 h of cultivation, dry cell biomass content measurement was done basing on cell dry weight (CDW) measurement, by taking 5 ml from each flask. The biomass centrifuged, washed, and vortexed with Mill-Q water two times, then the solids dried on weighted aluminum pans in oven running at 70°C for 24 hrs.
3.4 Wheat straw slurry solids and ash measurements
12
of continuous substrate feeding during enzymatic hydrolysis, the wheat straw slurry was diluted to 1/8 of its original concentration with Mill-Q water. Three part of slurry 200 g, 100 g, and 30 g was taken and diluted to 1/8 dilution with Mill-Q water. One part containing 30 g of the slurry diluted to 1/8 was sieved using an ordinary kitchen sieve. The solid fraction was then added into a beaker filled with 240 ml in order to have the same volume as the sieved liquid fraction. Because the wheat straw was pretreated in the acidic condition, the initial pH 1/8 Diluted slurry, solid fraction and the sieved liquid fraction was 2.83, 2.88 and 2.91 respectively taken using a pH meter. However, the cellulase enzyme function optimally at pH 5-5.5 (Ishola et al. 2015); therefore the pH of the slurry has been adjusted by adding 10M NaOH solution to about 5. Solids and ash measurements were determined by following NREL protocol (Sluiter et al. 2012a). Suspended solids and total solids content were determined by sampling 5 ml from each of three phases. For the SS, the sample was centrifuged at 3000xg for 2 min, removing the supernatant; the remaining part was washed with Mill-Q water and vortexed for two times. The washed solids were then poured on weighted aluminum pans and dried together with aluminum pans for containing sample for total solids measurement in the oven at 70°C for 24 hours. For ash measurement, the 5 ml for each fraction and poured in weighted ceramic crucibles and kept in the furnace at 550°C for 3 hours.
3.5 Enzymatic hydrolysis of the wheat straw slurry
The substrate was enzymatically hydrolyzed in shaking flask employing different temperatures and enzyme loadings to optimize the hydrolysis conditions. The hydrolysis experiments were performed in 250 ml Erlenmeyer flasks placed in different shaking water bath running at 30°C, 35°C and 50°C for 48 hours, the final volume was 100 ml for all assays. The cellulase activity was determined following the filter paper units (FPU) methodology according to Adney (2008). Enzymatic hydrolysis experiments were carried out in duplicate to determine the best condition process in terms of enzymes and temperature and to attain a medium with suitable sugar contents for bioethanol production. Table 2 represents a summary of all enzymatic hydrolysis tests performed with enzyme content employed. Considering the activity of cellulose to be around 130 FPU/ml, required amount of enzyme (E) was calculated as follow:
SS content of whole stillage (g/L)*volume of the medium (l) (0.1l= g SS in 100 ml of the medium FPU = x FPU/g SS* g SS in medium
13 Table 2. Enzymatic hydrolysis tests performed
Number of Flasks containing media SS in 100 ml of media X FPU/g SS FPU Enzyme (ml) Temperature condition °C
Flasks of 1/8 D. slurry media 2.050 8.7 17.831 0.137 35
Flasks of 1/8 D. slurry media 2.050 8.7 17.831 0.137 50
Flasks of 1/8 D. slurry media 2.050 12.1 24.799 0.191 35
Flasks of 1/8 D. slurry media 2.050 12.1 24.799 0.191 50
Flasks of 1/8 D. slurry media 2.050 15.6 31.972 0.246 35
Flasks of 1/8 D. slurry media 2.050 15.6 31.972 0.246 50
Flasks of 1/8 D. solid fraction 0.157 12.1 1.896 0.015 50
Flasks of 1/8 D. liquid fraction 1.509 12.1 18.259 0.140 50
Flasks of 1/8 D. slurry media 2.050 0 0.000 0.000 50
Different samplings were taken in duplicate at 0, 6, 30, 24, and 48 h. The aliquot of each sample was centrifuged at 15000×g for 2 min in which the supernatant was analyzed in HPLC to determine the content of sugars especially glucose and xylose.
3.6 Enzymatic hydrolysis and yeast cultivation in shake flasks
Simultaneous saccharification and fermentation (SSF) and separate saccharification and fermentation (SHF) experiments were performed in 250 ml Erlenmeyer flasks. 125 g and 100 g portion of wheat straw slurry were each 1/8 diluted with Mill-Q water and mixed well with a magnetic stirrer. The initial pH of the slurry was measured to be around 2.80, adjusted to pH 5 with 10M NaOH. The suspended solid measurements were done following the same procedure in section (3.2). The medium for SSF and SHF were prepared by adding 100 ml diluted slurry in 10 flasks of 250 ml Erlenmeyer flasks and autoclaved at 120°C for 20 min. Erlenmeyer flasks were placed in two different shaking water baths at 120 rpm one at 30°C and another at 50°C for 48 h. All experiments were carried out in duplicate for 8.7, 12.1 and 15.6 FPU/g SS at 30°C for enzymatic hydrolysis, 12.1 FPU/g SS for SSF at 30°C and 12.1 FPU/g SS at 50°C and 30°C for SHF, all for 48 h. The medium for SHF which was hydrolyzed at 30°C and 50°C after 48h was loop inoculated with S. cerevisiae and kept in shaking water bath running at 120 rpm at 30°C for 48 h. The amount of required enzyme considering the suspended solid (SS) were calculated following the same procedure in section (3.3) and are presented in the Table3. Sampling was done in duplicates at 0, 6, 30, 24, and 48 h for all experiments. An aliquot of each sample was centrifuged at 15000xg for 2 min. The supernatant was analyzed by HPLC to determine the change in sugars, ethanol, and inhibitor.
14 Number of Flasks containing media SS Content/ (g/l) SS (100ml media) X FPU/gSS FPU Enzyme (ml) Temperature condition°C Processes Flasks of 1/8 D. slurry media /125g 23.473 2.347 8.7 20.422 0.157 30 Enz. Hydrolysis Flasks of 1/8 D. slurry media /125g 23.473 2.347 12.1 28.403 0.218 30 Flasks of 1/8 D. slurry media /125g 23.473 2.347 15.6 36.618 0.282 30 Flasks of 1/8 D. slurry media /100g 23.430 2.343 12.1 28.350 0.218 30 for SSF Flasks of 1/8 D. slurry media /100g 23.430 2.343 12.1 28.350 0.218 30 For SHF Flasks of 1/8 D. slurry media /100g 23.430 2.343 12.1 28.350 0.218 50 For SHF
3.7 Membrane bioreactor design, set-up and defining filtration parameters
3.7.1 Membrane Bioreactor set-up
The second generations integrated permeate channels (IPC) membrane modules were used during experimental work supported by double filtration layers placed in 3D spacer-fabric support manufactured and supplied by the Flemish Institute of Technological Research (VITO NV, Belgium). The IPC membrane modules used are hydrophilic made by polyethersulfone (PES) materials with a pore size of 0.3 µm for microfiltration and the surface area 0.01372 m2 (0.00686 m2 for each panel). The second generation IPC membrane panels contain high quality of resisting the high-pressure differences during filtration and backwashing, which is an advantage to use the as submerged MBR. Two membrane panels were placed in parallel in spacer-box and kept inside 4.0 l Belach WebAnt® reactors (Belach Bioteknik AB, Skogås, Sweden) with 2.5 L as the working volume for all experiment. Actually, the ethanol production process needs more attention in order to avoid any contamination. Therefore, all materials and equipment were sterilized and disinfected before use. The 4.0 l bioreactors were autoclaved together with tubes at 121°C for 20 minutes. But, PC membrane panels cannot resist higher temperature, so they were chemically cleaned and disinfected before each experiment. Firstly, 2.5 l of 2% sodium hydroxide (NaOH) was transferred into a bioreactor for 45 min-1hour at 45°C, then after the bioreactor was two times drained and rinsed with sterile Mill-Q water. Secondly, 2.5 l of 1% phosphoric acid (H3PO4) was added into
15
cleaning procedures were performed before each membrane bioreactor experiment for all processes.
3.7.2 Pure water, wheat straw slurry and hydrolysate filterability measurements
Second generation IPC membrane panels used in this experiment for either enzymatic hydrolysis or fermentation can have different operating parameter depending on kind of liquid or solution with different viscosity applied to them. In this experiment parameters such as transmembrane pressure (TMP), flux and permeability were defined comparing membrane filtration performance with pure water, diluted wheat straw slurry and pure water after wheat straw slurry filtration (figure 3).
The membrane filtration performances were first tested with pure water using microfiltration IPC membrane panels placed in 4.0 l Belach WebAnt® reactors (Belach Bioteknik AB, Skogås, Sweden), cleaned and disinfected by the process explained in the (section 3.7.1). The temperature and nitrogen gas flow rate, the liquid level in the reactor of 35°C, 3 to 4 l/min and 2.5 l respectively were constantly controlled by WebAnt® controlling unit (Belach Bioteknik AB, Skogås, Sweden). The experiments were done comparing different feeding flow rates and permeate flow rates. Watson-Marlow 403/R1 peristaltic pumps (Watson Marlow, United Kingdom) were used to pump in the substrate to the membrane bioreactor and to pump out permeate to the permeate tank when required transmembrane pressure (TMP) was developed to the permeate side of the membrane panels. The permeate before being transferred into the permeated tank, they passed through a 710 Atrato ultrasonic flowmeter (Titan Enterprises Ltd., United Kingdom) and a pressure sensor PMC131 (Endress+Hauser AB, Solna, Sweden) in order to measure the permeate flow rate and the pressure on the permeate line respectively. The pressure sensor reader and flowmeters were connected to a computer and received data logged for further analysis. In order to have an adequate mixing of the high SS medium for suitable mass transfer as well as enhancing in situ membranes cleaning considerable amounts of nitrogen gas supplied as air/gas sparging. All experimental tests were performed in duplicate. The method used to measure the amounts of suspended solids (SS) as well
16
as the enzyme needed for 12.1 FPU/g SS are well described in sections(3.4) and (3.5) respectively. The initial pH of diluted slurry was 2.82 and was adjusted to pH 5 with 10M NaOH. The operating parameters for the pure water filtration and diluted wheat straw slurry are described in the table below and following paragraph:
Table 4. Different parameters followed during membrane filtration using pure water and diluted wheat straw slurry
The pure water permeability tests were comparing before and after slurry filtration for one hour. All samplings, feeding tank refilling and permeate tank emptying were performed in aseptic condition. Sampling was done in duplicate at 0, 3, 18, 21, 24, 27, 42, 45 and 48 hours where of 3 ml of each was taken and centrifuged, then the supernatant was kept into the friseur for later HPLC and while the solids part used for change in suspended solids measurements. 10 ml was taken in 13 ml tubes for later change in viscosity measurement.
Pure water replicate 1 1/8 Diluted wheat straw slurry replicate 1 Feed Flow rate of 10% with 8.1rpm and permeate
flow of 10% with 6.1rpm.
Feed Flow rate of 15% with 13.4rpm and permeate flow of 15% with 10.3rpm.
Feed Flow rate of 20% with 18.7rpm and permeate flow rate of 20% with 14.5 rpm
Feed Flow rate of 25% with 24.0rpm and permeate flow rate of 25% with 18.7rpm
Feed flow rate of 30% with 29.2rpm and permeate flow rate of 30% with 22.9rpm
Feed flow rate of 35% with 34.5rpm and permeate flow rate of 35% with 27.2rpm
Backwashing flow rate 0.6l/h Nitrogen gas flow rate 3.5l/min
Process cycle of 4.5min forward flow and 0.5 min backwashing
Temperature 35°C
Double Membrane with pore size of 0.3µm each Working volume 2.5l
Sampling time 15 second
Process running for 1hour for each steps
Feeding flow rate of 20% with 18.7rpm Permeate flow rate of 0.3l/h
Backwashing flow rate 0.6l/h Nitrogen gas flow rate 3.5l/min Temperature 35°C
Process cycle 4.5min forward flow and 0.5 backwashing
Sampling time 15 second Working volume 2.5l
Double MF Membrane with pore size of 0.3µm each
Used 12.1FPU/gSS for enzymatic hydrolysis
Fatty acid ester antifoam 200µl at the first start.
17
3.7.3 Yeast cultivation on semi-synthetic media and wheat straw hydrolysate to determine optimal operating condition
The yeast cultivation in different semi-synthetic medium was completed following methods described in section (3.3). The yeast was used to cultivate the wheat straw hydrolysate as permeate from membrane filtration and enzymatic hydrolysis of 1/8 diluted wheat straw slurry. Fortunately, the yeast consumed both glucose and xylose in semi-synthetic media as it described in section (4.5), on the other hand showed resistance to xylose consumption in wheat straw hydrolysate, because of lack of nutrients source in hydrolysate. Consequently, seven different medium cultivations were done to both semisynthetic media and wheat straw hydrolysate in order to figure out the optimal and good combination for next batch and continuous fermentation. The table5 illustrate the different materials contained in each medium for both semisynthetic media and wheat straw hydrolysate:
Table 5. The medium composition (Med.1. to 7.) of semi-synthetic media and wheat straw hydrolysate
Semi-synthetic media Wheat straw hydrolysate
Med. 1. 7.5g/l glucose+ 6.5g/l Xylose
Med.2. 7.5g/l glucose+ 6.5g/l Xylose + 2.5g/l yeast extract+ 2.5g/l peptone
Med. 3. 7.5g/l glucose+ 6.5g/l Xylose + 5g/l (NH4)2SO4
Med. 4. 7.5g/l glucose+ 6.5g/l Xylose + 2.5g/l yeast extract+
2.5g/l (NH4)2SO4
Med. 5. 7.5g/l glucose+ 6.5g/l Xylose + 5g/l yeast extract+
2.5g/l KH2PO4
Med. 6. 7.5g/l glucose+ 6.5g/l Xylose+ 5g/l
(NH4)2SO4+2.5g/l KH2PO4
Med.7. 7.5g/l glucose+ 6.5g/l Xylose +2.5g/l KH2PO4
Med. 1. Hydrolysate
Med. 2. Hydrolysate+ 2.5g/l yeast extract +2.5g/l peptone
Med. 3. Hydrolysate+ 5g/l (NH4)2SO4
Med. 4. Hydrolysate+ 2.5g/l yeast extract+ 2.5g/l
(NH4)2SO4
Med. 5. Hydrolysate+ 5g/l yeast extract+ 2.5g/l
KH2PO4
Med. 6. 7.5g/l glucose+ 6.5g/l Xylose+ 5g/l
(NH4)2SO4+2.5g/l KH2PO4
Med. 7. 7.5g/l glucose+ 6.5g/l Xylose +2.5g/l
KH2PO4
18
analysis was done at 1,18,24,42 and 48 h and pH measurement at 0, 24 and 48 h. Dry cell mass measurement was also done after 48 h.
3.8 Double stage membrane bioreactors for continuous enzymatic hydrolysis and fermentation
3.8.1 Pure water filterability to define filtration parameters
Different MBR filtration trials were performed using pure water by comparing various permeate flux of 21.9 LMH, 29.1 LMH, 36.4 LMH, and 43.7 LMH with sampling time from10 to 50 sec and relaxation time of 8 sec or 6 sec after 30 sec of backwashing. Ethanol evaporation analysis was done by adding 5.5, 11 and 30 g/l of 99.9% ethanol into a reactor filled with 2.5 l of pure water. Parameters taken into accounts were 30°C and 35°C temperatures each with gas/air flow rate 3 l/min and 4 l/min for 5 h considering the fact that the MBR filtration was carried out in a continuous process.
3.8.2 Wheat straw slurry dilution and MBR assisted enzymatic hydrolysis in batch and continuous process
19
sparging with a flow rate of 0.4l/h was provided to enhance perfect mixing inside MBR. Starting with hydrolysate permeates liquid containing fermentable sugars and inhibitors, to a certain amount of 400 ml in the buffer tank, was used as substrate feed-in continuous fermentation. The same processes were followed to all applied fluxes 21.9, 36.4 and 51 LMH.
3.8.3 Yeast pre-culture medium preparation
To prepare required yeast inoculum for continuous fermentation in MBR, yeast pre-cultures were prepared in 250 ml Erlenmeyer flasks containing 100ml of yeast extract peptone dextrose (YPD) broth. During continuous fermentation cell conversion rate depend on the initial amount of cells. Thus different pre-cultures were employed to optimized better performance. Low concentration pre-cultures broth containing 7.5 g/l glucose, 6.5 g/l xylose, and 5 g/l yeast extract were prepared. High concentration pre-cultures medium, as well as fermenter media contents, were chosen to referring to the research done by Mahboubi et al. (2017a) in which pretreated wheat straw hydrolysate was used as substrate but with different operating conditions comparing to this experiment. The substrate was 1/4 diluted while in this experiment was 1/8 diluted. In this case, to prepare the pre-culture and fermenter media during the experiment, the same compounds and half of the concentration were taken into account. The high pre-culture concentration broth containing 12.5 g/l glucose, 12.5 g/l xylose, 10 g/l peptone, and 5 g/l yeast extract were prepared. The pre-cultures were loop inoculated and placed in a shaking water bath (Grant OLS 200, Grant instrument ltd, UK) at 30°C and 120 rpm in 24h for low concentration broth and 48 h for high concentration broth. After 24h and 48h of incubation, biomass content of cultures of 2.16±0.005 g/l and 6.56±0.3 g/l respectively were used to inoculate the MBR for fermentation.
3.8.4 MBR assisted fermentation batch process to optimize operating condition
During the batch process, the amounts, contents of batch medium and initial inoculum were prepared relating to the pre-culture prepared in (section 3.8.3). Low inoculum of 200 ml was used to fermenter filled with inoculating 2.3 l autoclaved broth containing 7.5 g/glucose, 6.5 g/l xylose, 5 g/l yeast extract and 2.5 g/l of KH2PO4. On the other side, a high inoculum of 300 ml was used
to inoculate fermenter filled with 2.2 L autoclaved broth containing 25 g/l glucose, 25g/l xylose, 5 g/l peptone, 2.5 g/l yeast extract, and 2.375 g/l KH2PO4 (figure 4). During batch fermentation,
20
air sparging with a flow rate of 0.4l/min was supplied to provide minimal aeration for cell growth and medium mixing.
Figure 4. Inoculation and batch process in MBR
3.8.5 Continuous fermentation of wheat straw hydrolysate
Hydrolysate, filtrate rich in fermentable sugars (glucose and xylose) filtrated into the buffer tank, was continuously fed into fermentation MBR. The fermentation was performed anaerobically at the temperature of 30°C while the nitrogen gas with a flow rate of 0.4l/min was provided to have a better mixing inside the fermenter. The permeates were continuously filtrated into a permeate tanks comparing different constant flow rates of 21.9 LMH, 36.4 LMH, and 51 LMH together with backwashing flow rates of 0.6 l/h, 1l/h, and 0.14l/h respectively (figure 5). Throughout the continuous processes, Watson-Marlow 403/R1 peristaltic pumps (Watson Marlow, United Kingdom) were used for feeding or permeate removal. Moreover, liquid flowrates, pressure on the main reactors and permeate lines were provided to Mefias® by 710 Atrato ultrasonic flowmeters (Titan Enterprises Ltd., United Kingdom) and PMC131 pressure sensors (Endress+Hauser AB, Solna, Sweden), respectively. In order to compensate the lack of nutrients source in hydrolysate, 10ml solutions rich in nutrients transited to2.5 g/l of yeast extract, 2.5 g/l of (NH4)2SO4 and 3.5g/l
21
bioreactors, pumps involved in feeding and permeate removal, pressure sensors and flow meters were completely synchronized and automated, using the LabView-based Mefias® software specifically customized and developed by VITO for this research work. During continuous fermentation, different samples were taken every 8 and 16 h to monitor sugar consumption and metabolites production.
Figure 5. Double stage membrane based continuous hydrolysis and continuous fermentation
3.8.6 Membrane fouling measurement based on solid loading, backwashing and Flux
In order to check the extent and reversibility of membrane fouling, filtration of slurry and a hydrolyzed slurry of 1/8 and 1/2 diluted concentrations at different permeate flux were conducted. The medium was filtered at a low starting permeate flux of 21.9 LMH for certain filtration time of 1 h then the filtration rate was raised to 36.4 LMH held for 1 h and then back to the initial flow 21.9 LMH for 30 min, followed by 51 LMH for 1 h then back again to 21.9 LMH flow for 30 min. This trend was repeated stepwise for flow rates with and without backwashing (BW).
Feed tank MBR-Hydrolysis
22 - Runs performed at 35°C (initial pH 5)
- Permeate recirculation
- Clean water filtration before and after the run was considered with the same stepwise regime.
The changes of TMP, total resistance of membrane and cake layer resistance for all fluxes with and without backwashing were taken into consideration. The amounts of total resistance and cake layer resistance are calculated according to equations 1 and 2 respectively.
𝑱 =𝑻𝑴𝑷
µ𝑹𝑻
(1)
𝑹𝑻= 𝑹𝒎+ 𝑹𝑪 + 𝑹𝒇 (2)
Where J is permeate flux, µ is permeate viscosity, RT is total resistance, Rm membrane resistance
(clean water filtration), RC is the cake layer resistance and Rf is irreversible fouling resistance. As
the sugar content (Brix degree) of the hydrolysate is very low to make perceptible change in medium viscosity and as the hydrolysis temperature is 35°C, the µ of water at 35°C (0.789×10-3
Pa.s) has been considered for the permeate. As in microfiltration irreversible fouling due to particle penetration in the membrane pores rarely occurs resulting in considerably smaller Rf resistance
than Rc, Rc is used to represent resistance due to cake layer and irreversible fouling (Choo et al. 1996).
3.9 Analytical methods
High performance liquid chromatography (HPLC) (Walters 2695, Walters Corporation, and Milford, USA) was used to determine and analyze the concentration of different compounds and metabolites in cultivation medium as well as in hydrolysate. During analysis, a hydrogen based ion-exchange column (Aminex HPX-87H, Bio-Rad, Hercules, USA) working at 60°C with 5 mM H2SO4 eluent flowing 0.6ml.min-1 was used to identify and quantify sugars, inhibitory compounds
23
Aluminum pans then drying the washed cell in 70°C oven for 24 hrs. Acid insoluble lignin, acid soluble lignin as well as carbohydrate and ash content from solid retained during continuous membrane filtration assisted hydrolysis were estimated using the procedure Determination of Structural Carbohydrates and lignin in biomass (Sluiter et al. 2012b).
24
4. Results and discussion
The main goal of this project was to integrate MBR into the lignocellulosic bioethanol production process in terms of process intensification. The performance of microfiltration membrane in such a process was also investigated. Different parameters and conditions were investigated to set up the system. Membrane filtration performance was firstly investigated using clean water and hydrolysate. Continuous membrane assisted enzymatic hydrolysis and fermentation in a controlled interconnected double-staged submerged MBRs were investigated by comparing different constant permeate fluxes of 21.9 LMH, 36.4 LMH, and 51 LMH. Note that, both hydrolysis and fermentation were carried out in parallel in separate submerged MBRs. Results conditions optimization, the setup processes by means of double-stage submerged MBR are well described in the following sections.
4.1 Ethanol production in semi-synthetic media
25
Figure 6. Changes in the concentration of metabolites during cultivation in semi-synthetic media at 30°C.
Glycerol formation was relatively low but slightly increased to 0.58±0.12 g/l at 30 h and remains almost the same till 48 h. Glycerol is synthesized by yeast for equilibrating the intracellular redox balance by converting the surplus of NADH generated during biomass formation to NAD+(Carlos et al. 2002). Xylitol formation was totally low as 0.22±0.25 g/l of xylitol was only formed in 48 h. One of the aims of this project is to achieve the highest yeast co-utilization of glucose and xylose in a continuous MBR assisted fermentation. Although the slow consumption of xylose was noticed compared to glucose consumption, different inoculum and condition optimization were investigated (section 4.4).
4.2 Enzymatic hydrolysis of the wheat straw slurry, liquid and solid fractions
A number of enzymatic hydrolysis experiments were carried out to determine the best condition in terms of enzyme loading and temperature to attain a medium with a suitable sugar contents for bioethanol production. Throughout the study, the 1/8 diluted slurry is regarded as a slurry to skip repetitions and writing complications. Whole slurry and solid and liquid fractions of slurry after sieving using an ordinary kitchen sieve were enzymatically hydrolyzed well described in (section 3.5). The suspended solid contents in the diluted slurry, solid and liquid fractions were 20.49±0.02 g/l, 1.56±0.41 g/l, and 15.09±0.12 g/l respectively.
Figure 7a and b presents the concentration of sugar released during enzymatic hydrolysis at enzyme loadings of 8.7, 12.1, and 15.6 FPU/gSS at 30, 35 and 50°C. The concentrations of glucose and xylose released at 50°C after 48h were high compared to sugars released at 30 and 35° C. Considering the totals sugar released during enzymatic hydrolysis, 50°C condition seems to be the
0 5 10 15 20 25 0 24 48 C o n ce n tra ti o n in g/l Time in (h)
26
27
Figure7. Changes in the concentrations of sugars during enzymatic hydrolysis of the 1/8 diluted wheat straw slurry at 30°C, 35°C and 50°C, a) glucose concentration for 8.7, 12.1, and 15.6 FPU/g SS and, b) xylose concentration for 8.7, 12.1, and 15.6 FPU/g SS.
0 2 4 6 8 10 12 14 16 0 24 48 C o n ce n tra ti o n in g/l 0 1 2 3 4 5 6 7 8 0 24 48 C o n ce n tra ti o n in g/l Time (h)
28
Figure 8 represents the changes in the concentration of glucose and xylose as a result of enzymatic hydrolysis of sieved liquid and solid fraction of wheat straw at 50°C with 12.1 FPU/g SS. The concentration of glucose and xylose in sieved liquid fraction of 9.90±0.19 g/l and 6.79±0.007 g/l, respectively, was higher than that of the solid fraction (remaining from sieving) of 0.50±0.03 g/l and 0.04±0.004 g/l respectively. Lignin residues released during the pre-treatment of lignocellulose materials, increase the viscosity of the slurry and result not only in improper medium mixing and mass transfer but also in influencing cake layer formation which lead to membrane fouling (Mahboubi et al. 2017b). Hence, optimum solid loading is necessary in order to successfully apply membrane for separation purposes in lignocellulosic ethanol production. Consequently, the results of enzymatic hydrolysis of sieved liquid and solid fractions of slurry shows that if the slurry is first sieved before hydrolysis only negligible amount of sugars lost. However, slurry sieving benefit feeding and filtration processes. Figure 9 present the Composition and concentration of saccharides of liquid fraction of 1/8 diluted slurry and 1/8 diluted slurry enzymatically hydrolyzed at different temperature and with different amount of enzyme loadings.
Figure8. HPLC analysis results of enzymatic hydrolysis of 1/8 diluted sieved liquid and solid fractions of wheat straw slurry: (a) glucose concentration and (b) xylose concentration (50°C, 12.1 FPU/g SS). 0 2 4 6 8 10 12 0 24 48 C o n ce n tra ti o n in g/l Time (h)
Sieved liquid Solid fractions
(a) 0 1 2 3 4 5 6 7 8 0 24 48 C o n ce n tra ti o n in g/l Time (h)
Sieved liquid Solid fractions
29
Figure9. Composition and concentration of saccharides of 1/8 diluted slurry enzymatically hydrolyzed at different temperature and enzyme loadings.
0 2 4 6 8 10 12 14 16
Slurry 30°C 35°C 50°C Slurry 30°C 35°C 50°C Slurry 30°C 35°C 50°C
8.7FPU/gSS 12.1FPU/gSS 15.6FPU/gSS
C o n ce n tra ti o n (g/l )
30
4.3 SSF and SHF enzymatic hydrolysis and yeast cultivation
31
has proved to be effective and has been chosen for continuous double-staged membrane filtration assisted enzymatic hydrolysis and fermentation.
Figure 10. Changes in concentration of substrates and metabolites during simultaneous saccharification and fermentation (SSF) and separate hydrolysis and fermentation (SHF) performed in shake flasks for 48h: a) SSF at 30°C, b) Enzymatic hydrolysis and fermentation both at 30°C and c) Enzymatic hydrolysis at 50°C and fermentation performed at 30°C.
0 2 4 6 8 10 0 24 48 C o m ce n tra ti o n in g/l Time (h)
Glucose Xylose Ethanol Glycerol
(a) 0 2 4 6 8 10 12 14 0 24 48 72 96 C o n ce n tra ti o n (g/l ) Time (h)
Glucose Xylose Ethanol Glycerol
(b) 0 2 4 6 8 10 12 14 16 18 0 24 48 72 96 C o ce n tra ti o n (g/l ) Time (h)
Glucose Xylose Ethanol Glycerol
32
4.4 Effect of yeast cultivation on semi-synthetic media and wheat straw hydrolysate
33
Figure 11. Yeast cultivation of: a) semi-synthetic media and b) wheat straw hydrolysate and c) Change of inhibitors during yeast cultivation in hydrolysate.
0 5 10 15 20 25 0 24 48 72 C o n cn et ra ti o n (g/l ) Time (h)
Ethanol Glucose Glycerol Xylose Xylitol (a) 0 1 2 3 4 5 6 7 8 0 24 48 72 C o n ce n tra ti o n (g/l ) Time (h)
Ethanol Glucose Glycerol Xylose Xylitol
(b) 0 1 2 3 4 5 0 0.2 0.4 0.6 0.8 1 0 24 48 72 A ce ti c ac id (g/l ) Fu rf u ra l a n d H M F (g/l ) Time (h)
Furfural HMF Acetic acid
34
4.5 Optimization of membrane filtration parameters
4.5.1 Comparison of clean water and hydrolysate filterability
35
Figure 12. Comparison of (a) permeability and (b) TMP during membrane filtration of pure water, hydrolysate and pure water after hydrolysate filtration.
(a)
36
4.5.2 Comparison of constant fluxes to optimize interconnected membrane filtration performance
37
Figure 13. Pure water filterability comparing different permeate fluxes for interconnected double-staged submerged MBR (MBR-1 and MBR-2): (a) Flux; (b) TMP (c) Permeability
21.9LMH
36.4LMH 29.1LMH
38
4.5.3 Yeast growth optimization for batch and continuous fermentation
39
Figure 14. Yeast cultivation to in semi-synthetic media and wheat straw hydroysate in 7 different medium ( Med 1 to 7); (a), (b) and (c) semi-synthetic media and (d), (e) and (f) wheat straw hydrolysate. 0 1 2 3 4 5 6 7 0 24 48 C o n ce n tra ti o n (g/ l) Time (h)
Ethanol
(a) 0 1 2 3 4 5 6 7 0 24 48 C o n ce n tra ti o n (g/l ) Time (h)Ethanol
(d) 0 1 2 3 4 5 6 7 8 9 0 24 48 C o n ce n tra ti o n (g/l ) Time (h)Glucose
(b) 0 1 2 3 4 5 6 7 8 9 0 24 48 C o n ce n tra ti o n (g/l 9 Time(h)Glucose
(e) 0 2 4 6 8 0 24 48 C o n ce n tra ti o n (g/l 9 Time (h)Xylose
Med 1 Med 2 Med 3 Med 4
Med 5 Med 6 Med 7
0 1 2 3 4 5 6 0 24 48 C o n ce n tra ti o n (g/l ) Time (h)
Xylose
Med 1 Med 2 Med 3 Med 4
40 4.5.4 Ethanol evaporation measurement in MBR
In submerged MBRs process gas sparging /air scouring is one of method used to provide a better medium mixing and to mitigate membrane fouling rates mostly in anaerobic processes. However, the air/gas sparging rates in anaerobic submerged MBRs when used in ethanol fermentation may be the source of considerable amount of ethanol losses through excessive evaporation over time (Mahboubi et al. 2017a). In order to measure maximum amount of ethanol evaporated, the ethanol evaporation analysis was done before fermentation process at 35°C and nitrogen gas sparging of 0.4 l/h, by comparing initial ethanol concentration of 11 g/l and 5.5 g/l, the results are presented in figure15. The experiment was performed for 6 h and it was observed that 0.30 g/l and 0.21 g/l.h of ethanol were evaporated with high and low initial ethanol concentration respectively (figure 15). Therefore, as the MBR fermentation and hydrolysis were done at 30°C and 35°C respectively with air/gas flow rate of 0.4 l/h and considering the fact that the concentration of ethanol produced during fermentation was between 5-10 g/l, during results analysis, the amount of ethanol evaporated of 0.21 g/l.h was taken into account in all calculations.
Figure 15. Ethanol evaporation analysis in MBR at temperature of 35°C and air/gas flow rate of 0.4l/min comparing different initial ethanol concentration
41 4.6 Double-stage membrane biorector
The main goal of this part of the project was to investigate the performance of interconnected double-stage submerged membrane bioreactors (MBRs) assisted hydrolysis and fermentation in terms of process intensification. The processes were carried out comparing filtration performance during hydrolysis and fermentation at different constant permeate fluxes of 21.9 LMH, 36.4 LMH, and 51 LMH. Membrane assisted continuous hydrolysis interconnected together with membrane assisted continuous fermentation in which double IPC membrane supported in spacer boxes (well described in section 3.7.1) were submerged inside the 4 l reactor in which 2.5 l was working volume as illustrated in figure 16. After the batch processes, hydrolysate filtration from MBR based hydrolysis was started by feeding hydrolysate in a buffer tank reach out to a minimal volume of 400 ml and serve as initial feed in the fermenter which took around 2 hours. An integrated control system was applied, which combine both filtration and biological processes. Integrated control systems are the most promising on MBR control, they are most important due to their ability to reduce membrane fouling as well as sustaining sufficient biological/chemical removal (Ferrero et al. 2012).
42
The results showing the performance of membrane filtration during hydrolysis and fermentation are described in the following pages.
Figure 16. Experimental sketch (a) Automated sketch of the process from computer control and (b) Actual experimental set up.
43
4.6.1 Double-staged submerged MBR performance at constant flux of 21.9LMH
To investigate the effect of suspended solids and cell biomass accumulations on membrane, the filtration performed first at constant permeate flux of 21.9 LMH. The performance of membrane was defined considering the changes in membrane resistance and TMP, which depend on the cake layer formation on the membrane surface, resulted mostly in membrane fouling. Generally, TMP supposed to remain constant or slowly increases when membrane performed at low fluxes. Even though the filtration happen in potentially fouling feed solution, the TMP and membrane resistance increase proportionally (Miller et al. 2014). During filtration flux of 21.9 LMH the resistance and TMP increased slowly, at constant rate with time and after 90h, the trends remained stable till the end of filtration 264 h.
Considering initial TMP of 0.007 bar during continuous hydrolysate filtration, a small increase of TMP with negligible increase rate of 0.0005 bar/day was noted. Even though the concentration of suspended solids (SS) increased at the rate of 8.5 g/l.day between 0 to 142 h reached 57.5±0.05 g/l, to maintain stable SS content in reactor, draining was applied per day by removing 340 ml from reactor. The TMP showed a negligible increase and was gradually increased proportionally to the membrane resistance (figure 17b and c (red)). Therefore, during continuous hydrolysate filtration, membrane showed a great performance at TMP between 0.007 and 0.011 bars for a period of 11 days (264 h).
44
fermentation. Although membrane filtration at a constant flux of 21.9 LMH shows an effective performance during continuous hydrolysis and fermentation, filtration performance at higher fluxes were necessary and investigated considering membrane selectivity, ethanol productivity and membrane functioning in terms of membrane fouling tolerance.
Figure17. Membrane bioreactor filtration performance during enzymatic hydrolysis and fermentation with a constant permeate flux of 21.9LMH: (a) Permeate flux, (b) TMP, Suspended Solids (SS) and cell biomass (c) Membrane resistance, Suspended Solids (SS) and cell biomass.
(a)
45
4.6.2 Double-staged submerged MBR performance at constant flux of 36.4LMH
46
Figure 18. Membrane bioreactor filtration performance during enzymatic hydrolysis and fermentation with a constant permeate flux of 36.4LMH: (a) Permeate flux, (b) TMP, Suspended Solids (SS) and cell biomass (c) Membrane resistance, Suspended Solids (SS) and cell biomass.
47
4.6.3 Double-staged submerged MBR performance at constant flux of 51 LMH
This part of the experiment was aimed to study the effect of high flux on membrane selectivity, and membrane fouling tolerance and compare membrane filtration effectiveness at constant of 51 LMH 36.4 LMH, and 21.9 LMH. Figure 19 illustrates the results of membrane filtration performance at the constant flux of 51 LMH during continuous hydrolysis (green) and fermentation (blue). It was observed that the high accumulation rate of suspended solids of 26.75 g/l.day at constant filtration flux of 51 LMH, caused a rapid increase of materials deposited and membrane pores blockage, which resulted in rapid membrane fouling. However, it was indicated by a successive increase in membrane resistance which resulted in high TMP upturn with a high increasing rate of 0.035 bar per day increase from 0.016 bar up to TMP of 0.146 bar in only 52 h (figure 19b and c (green)). The TMP rises exponentially from the very beginning due to the presence of high flux. After 52 h, the system showed the needs of more maintenance and adjustment for feeding and filtration and chemical cleaning as backwashing (BW) has not been as effective. It was concluded that, during continuous filtration with high flux, membrane lose its capacity and lead to low permeability and high membrane resistance in short period of time. Although, the rate of suspended solid 26.75 g/l.day was high compared to the rate of cell biomass accumulation inside the reactor around 1.43 g/l.d, the final TMP during hydrolysate filtration was low and significantly different to the final TMP during bioethanol fermentation and filtration a P-value = 0.0001. During bioethanol fermentation and filtration, it was noticed that membrane performed at high TMP up to 0.34 bar in only 19 h and slowly increased to 0.4 bar till 52 h as it was presented in (figure 19b and c (blue)). In addition, statistical analysis of final TMP values for all applies fluxes during filtration assisted hydrolysis proved that there is a significant difference between final TMP with a P-value = 0.006, while no significant difference observed between final TMP comparing 36.4 LMH and 51 LMH with a P-value = 0.256.
48
Figure 19. Membrane bioreactor filtration performance during enzymatic hydrolysis (green) and fermentation (blue) with a constant permeate flux of 51 LMH: (a) Permeate flux, (b) TMP, Suspended Solids (SS) and cell biomass (c) Membrane resistance, Suspended Solids (SS) and cell biomass.
(a)
49
4.7 Continuous enzymatic hydrolysis and filtration of pretreated wheat straw slurry One of the projects’ goals was to ensure the effectiveness of enzymatic hydrolysis during continuous hydrolysate filtration at constant permeate flux of 21.9 LMH, 36.4 LMH and 51 LMH. Enzymatic hydrolysis of the wheat straw slurry was accomplished by adopting different enzyme loading to optimize the operating conditions. The experiments started by batch steps for 48 h and followed by continuous hydrolysis while hydrolysate was continuously filtered in the buffer tank by the help of double IPC microfiltration membrane. Initially, wheat straw slurry before use was 1/8 diluted, sieved with an ordinary kitchen sieve and autoclaved for 20 min at 121°C. The reasons of sieving were well explained in Mahboubi et al. (2017a) article. Though, continuous enzymatic hydrolysis and fermentation were carried in parallel into double stage MBR, where hydrolysate filtered in buffer tank was immediately fed into the fermenter. Batch hydrolysis and fermentation were completed in 48 h and 24 h respectively and after each batch phases, continuous feeding and filtration started at the same time. Batch hydrolysis started 24 h after pre-cultures incubation in shaking flasks, to complete 48 h of batch hydrolysis at the same time with 24 h of batch fermentation. The results are presented in the following paragraphs.
4.7.1 Continuous hydrolysis with low enzyme loading at constant permeate flux of 21.9LMH
50
filtrated into the buffer tank at the constant flux of 21.9 LMH. In this phase, enzymatic hydrolysis was completed in 240 h with 192 h of continuous hydrolysis.
51
Figure 20. Results from enzymatic hydrolysis and hydrolysate filtration at low enzyme loading and filtration permeate flux of 21.9LMH: (a) change in sugars (glucose and xylose), inhibitors, in MBR; (b) Amount of sugars (Glucose and glucose), and inhibitors filtered into buffer tank.
0 2 4 6 8 10 12 14 16 18 20 0 24 48 72 96 120 144 168 192 216 240 C onc ent ra ti on (g /l ) Time (h)
Acetic acid Glucose Xylose Furfural HMF
0 5 10 15 20 25 30 48 72 96 120 144 168 192 216 240 C on cent ra ti on ( g/ l) Time (h)
Acetic acid Glucose Xylose Furfural HMF
(b) (a) Continuous
52
4.7.2 Continuous hydrolysis with high enzyme loading at constant permeate flux of 21.9 LMH
